Trill makes note after play very quietly

• Mar 13, 2018 - 20:32

I came across this bug with trill playback. After playing back a trill, whether it is a line or ornament doesn't matter, the first note is very quiet. Playing back from the note after the trill, the sound is as expected, just as loud as the other notes around it. Reproducible on the latest nightlies, both 2.2 (OS: Windows 7, Arch.: i386, MuseScore version (32-bit): 2.2.0, revision: 86d468b) and master (OS: Windows 7 SP 1 (6.1), Arch.: i386, MuseScore version (32-bit): 3.0.0, revision: c08fe0d)

The bug seems to only affect the voice in which the trill is. To illustrate, I attach the score below. The problem occurs on the two Ds at the ends of bars 3 and 5. The other voices in bar 3 plays back normally. Playing back from either D, thus eliminating the playback of the preceding trill, makes the playback of the D noticeably louder. It also depends on which instrument is used how bad the problem is. Piano and harpsichord is particularly bad, while clarinet is much less noticeable.

If this has been reported, please point me to the discussion, because I did not find it.

Louis

Attachment Size
Trillerfouttoets.mscz 5.03 KB

Comments

• Mar 14, 2018 - 00:37

I can't reproduce. I'm on arch linux x86-64.

I actually think your brain might be fooling you into thinking that you don't hear the D after the trill, simply because you have heard so much trilling before it that you think that D belongs to the trill, when really that D sound does indeed belong to the subsequent note.

Reply

• Mar 14, 2018 - 01:19

For meas 26, I'm first deleting notes in the bass clef, so I can hear the treble clef better. Then I play that measure. I clearly hear the final pitch in that measure is B. But it sounds like that B is part of the trill. If I then delete that final B note, and play the measure, I hear that measure clearly ends on an A. So that means that the trill ends on pitch A. But because the next note of B could be potentially mis-interpreted as belonging to the B, your brain is fooling you into thinking that the B belongs to the trill.

But meas 29 is interesting, and you might be onto something. I've deleted everything except the treble clef of that measure (attached) and exported that wav (with the zita echo effect delayed) and opening it in audacity and played back at quarter the speed. And what I discover is interesting, because I don't clearly hear the final C be re-articulated, although I do see a little bump in the waveform. Tweestemmige_vinding_in_A_meas29.mscz

Analysing a bit further, I'm trying that measure 29 but now with the trill element only on the inital half note, and I hear something interesting: at the end of the measure, I hear both the C# and D sound at the same time like a chord: Tweestemmige_vinding_in_A_meas29_modified.mscz

So I need to think about this. Sounds like musescore is playing both the note after the trill and also the final trill note, and they overlap.

Reply

• Mar 14, 2018 - 01:52

Trill playback is interesting the more I look at it. If I have:

trill-tied-over.png

What I hear is a trill of short notes which ends on a sustained D, which is interesting because if the trill is not tied over:

trill-not-tied-over.png

it ends on a short D.

So the tie is somehow causing the final note of the trill to be sustained.

I'm not familar with how the trills are exactly implemented...but something seems strange about that.

Reply

Do you still have an unanswered question? Please log in first to post your question.